Pregnancy diagnosis in pigs: a field study comparing linear-array real-time ultrasound scanning and amplitude depth analysis.

Between days 24 and 32 after mating/insemination, 881 pigs (785 pregnant and 96 not-pregnant) were tested for pregnancy on a commercial farm with a linear-array real-time ultrasound scanner. 5-7 Days later, 785 of these animals (708 pregnant and 77 not-pregnant) were tested again with A-mode equipment by farm employees. Confirmation of pregnancy was based on recorded farrowings or abortions; confirmation of non-pregnancy was based on return to oestrus and rebreeding, recorded non-farrowing, or inspection of the uterus of culled animals at the slaughterhouse. From the number of correct positive (a), incorrect positive (b), correct negative (c) and incorrect negative (d) diagnoses, a sensitivity (a/a + d) of 100% versus 97.5%, a specificity (c/c + b) of 90.6 versus 55.8%, a positive predictive value (a:a + b) of 98.9% versus 95.3% and a negative predictive value (c:c + d) of 100% versus 70.5% were calculated for the real-time ultrasound technique versus A-mode technique. It was concluded that real-time ultrasound scanning provides a very accurate technique for pregnancy diagnosis in pigs, enabling immediate decision making on treatment or culling of animals diagnosed as non-pregnant.